Remote hosting of applications is becoming increasingly common, and offers several advantages to remote users of the hosted applications. In this regard, a data center consisting of one or more servers may host applications, which may be accessed and/or otherwise used by remote user sites (e.g., “spokes”). Such a data center may beneficially provide application hosting, data storage, and/or data backup services, thus reducing the need for user sites to locally maintain costly, and potentially sizeable, computing infrastructure. For example, a single data center may host applications that service multiple hospitals, medical clinics, and/or the like, thus reducing the burden for maintaining a full computing infrastructure at each hospital site. Accordingly, from the perspective of a customer of hosted applications, a substantial portion of the computing equipment and information technology management costs may be offloaded to a remote application host.
However, in many instances, multiple customers, or spoke sites, will use overlapping local subnet addresses. Accordingly, management of networks supporting remote hosting may require an added layer of complexity in the form of the use of network address translation (NAT) devices to support communication between a hub and the spoke sites and/or between spoke sites given the overlapping local subnet addresses of some spokes.